Expectation/placebo-related mechanisms and specific effects of therapies show additive effects, such that a therapy is less effective if the placebo component is absent. So far, the placebo component has been disrupted experimentally by using covert administrations of treatments. Here, we show for the first time that disruption of expectation/placebo-related analgesic mechanisms may occur in a clinical condition, Alzheimer’s disease (AD). In order to assess the placebo component of a therapy, we used the recently developed open–hidden paradigm. A local anesthetic was applied, either overtly or covertly, to the skin of AD patients to reduce burning pain after venipuncture. The placebo (psychological) component is represented by the difference between the analgesic effect after open (expected) and after hidden (unexpected) application. We correlated the placebo component with both cognitive status and functional connectivity among different brain regions. We found that AD patients with reduced Frontal Assessment Battery scores showed reduced placebo component of the analgesic treatment. We also found that the disruption of the placebo component occurred when reduced connectivity of the prefrontal lobes with the rest of the brain was present. Remarkably, the loss of these placebo-related mechanisms reduced treatment efficacy, such that a dose increase was necessary to produce adequate analgesia. These findings highlight the active role of cognition and prefrontal lobes in the therapeutic outcome and underscore the need of considering a possible revision of the therapeutic approach in Alzheimer patients in order to compensate for the loss of the endogenous expectation and placebo mechanisms.